General inhibition by transforming growth factor β1 of thyrotropin and cAMP responses in human thyroid cells in primary culture

Molecular and Cellular Endocrinology, 95 (1993) 13-21 0 1993 Elsevier Scientific Publishers Ireland, Ltd. 0303.7207/93/$06.00

13

MCE 03038

General inhibition by transforming growth factor pl of thyrotropin and CAMP responses in human thyroid cells in primary culture Martine

Taton,

Fransoise

Lamy, Pierre P. Roger and Jacques

E. Dumont

Institute of Interdisciplinary Research, School of Medicine, Free Unil~er.sityof Brussels, Campus Erasme, Brussels, Belgium (Received

Key words: Thyroid

(human

cell); Transforming

growth

22 March

factor

1993; accepted

pl; Thyrotropin;

16 April 1993)

CAMP

Summary

Transforming growth factor pl (TGFPl) mRNA has previously been identified in human thyroid cells and this agent has been shown to inhibit DNA synthesis in thyroid cells of some other species. In normal human thyroid cells in primary culture, TGFPl inhibited inconstantly the low basal DNA synthesis and strongly the stimulation of DNA synthesis by equine growth factor (EGF) and serum, and by thyroid-stimulating hormone (TSH) acting through CAMP. This inhibition, by TGFPl, of the TSH and CAMP-dependent DNA synthesis was associated with an inhibition of PCNA (proliferating cell nuclear antigen) synthesis. TGFPl almost completely abolished the CAMP induced stimulation of iodide uptake and thyroperoxidase synthesis. It thus, like EGF, also acts as a dedifferentiating agent. Investigation of the pattern of protein synthesis by two-dimensional gel electrophoresis revealed that while TGFPl, by itself, increased the synthesis of only one protein, a tropomyosin isoform, it inhibited most of the effects of CAMP on protein synthesis (3.5 out of 45 CAMP-regulated proteins were affected). It also reversed the effect of CAMP on the morphology of the thyrocytes. The fact that TGFPl did not affect the increase in CAMP provoked by TSH in human thyroid cells while inhibiting most of the effects of dibutyryl CAMP in these cells suggests an action at a step distal to CAMP generation. These data reemphasize the potentially major role of TGFPl as a local modulator of the thyroid, showing for the first time that it inhibits both the growth and differentiation effects of TSH and CAMP on human thyroid cells.

Introduction

TGFP is the most potent growth inhibitor known for a wide variety of cell types in vitro including selected cell types of mesenchymal and myeloid origin, as well as nearly all epithelial, lymphoid and endothelial cells (Barnard et al., 1990). Since expression of TGF/3 mRNA has been detected in many cell types, including some which are inhibited by it, it has been hypothesized that the polypeptide may mediate a negative autocrine loop limiting normal cell proliferation. Activation of the secreted latent TGFp molecule could constitute a major regulatory step in the control of its action on cells (Roberts and Sporn, 1988). The molecular mechanism by which TGFP inhibits cell growth is

Correspondence to: M. Taton, I.R.I.B.H.N., School of Medicine, Campus Erasme, building C, B-1070 Brussels, Belgium. Tel.: 32 2 555-41-40; Fax: 32 2 555-46-55.

not known but recent studies indicate that its effects on c-myc expression (Pietenpol et al., 1990) and on the phosphorylation of the ~105 RB protein (Laiho et al., 1990) may be central events in this process. On the other hand, although the effects of TGFP on cellular differentiation have been examined in several systems, they differ depending on the cell type or marker of differentiation being studied. For example, while TGFP markedly potentiates the ability of follicle-stimulating hormone (FSH) to stimulate estrogen secretion in granulosa cells, its effect on adrenocortical cells is just the opposite: it strongly antagonizes the ability of adrenocorticotropic hormone (ACTH) to stimulate the secretion of cortisol (Roberts and Sporn, 1988). TGFp mRNA has been detected in human thyroid follicular cells (Griibeck-Loebenstein et al., 1989) though immunoreactive TGFp was only found in malignant thyroid epithelial neoplasia (Jasani et al., 1990). TGFP itself has been shown to inhibit DNA synthesis

14

and proliferation in porcine thyrocytes (Tsushima et al., 1988; Winter et al., 1989) and in the rat thyroid cell line FRTLS (Morris et al., 1988; Colletta et al., 1989; Pang et al., 1992) and [ “Hlthymidine incorporation into DNA in human thyrocytes from nontoxic goiter and Graves’ disease (Gr~beck-~ebenstein et al., 1989). On the other hand, TGFP was reported to enhance (Morris et al., 1988) or decrease TSH-stimulated iodine trapping (Colletta et al., 1989; Pang et al., 1992) in FRTL-5 cells and to decrease it in porcine thyroid cells (Tsushima et al., 1988). Its effect on CAMP accumulation in response to TSH is also controversial (Tsushima et al., 1988; Winter et al., 1989; Morris et al., 1988; Colletta et al., 1989). In a previous work we have shown that in primary cultures of normal human thyroid cells TSH, acting via CAMP, stimulates growth and promotes the expression of the differentiated functions of these cells (iodide trapping and thyroperoxidase synthesis) (Roger et al., 1988; Lamy et al., 1990). In this study we show that TGFpI antagonizes the effects of TSH and dibutyryl CAMP on DNA synthesis and gene expression in human thyroid cells. Materials

and methods

Materials

Bovine TSH was purchased from Armour Pharmaceutical Co (Chicago, IL, USA), collagenase from Worthington Biochemical Co (Freehold, NJ, USA) and dispase from Boehringer Pharmaceutical (Mannheim, Germany). Fetal bovine serum was obtained from Gibco (Paisley, UK); deo~ribonuclease, bovine insulin, human transferrin and dibutyryl CAMP from Sigma (St Louis, MO, USA); TGFPl was from British Bio-Technology Ltd. (Oxford, UK); [methyl-3H]thymidine was obtained from the Radiochemical Center (Amersham, UK). Methods Primary cultures

Normal thyroid tissue was obtained from 3 normal subjects, killed in traffic accidents, at autopsy immediately after death following a protocol approved by the Ethics Committee of the Medical School, and from 3 adult patients undergoing surgery for a hypofunctioning solitary “cold” thyroid nodule. The thyroid cells were cultured as described previously (Roger et al., 1988) in the following mixture, which constitutes the control medium: Dulbecco’s Modified Eagle’s Medium/ Ham’s F12 medium/ MCDB104 medium (2: 1 : 1 vol/vol/vol~ containing 2 x 10m3 mol/l glutamine, supplemented by 5 mg/l bovine insulin, 1.25 mg/l human transferrin, 40 mg/l ascorbic acid, 500 mg/l bovine serum albumin and if indicated, 1% fetal bovine serum. Antibiotics (lo5 U/l penicillin, 100 mg/l streptomycin, and 2.5 mg/l amphotericin B) were also added. No systematic differ-

ences were observed in the properties of normal thyroid cells prepared from both sources in this and in other studies (Roger et al., 1988; Lamy et al., 1990). Proliferation assay

DNA synthesis was estimated by measuring the incorporation of [ “Hlthymidine into 10% trichloroacetic acid-precipitable material or by the frequency of ~3H]thymidine-labeled nuclei, as estimated by autoradiography. The cells in the Petri dishes were incubated for 24 h in the complete medium containing 3 X 10V5 mol/l thymidine, low4 mol/l deoxycytidine and 10 mCi/l [“Hlthymidine. After removal of the medium, the cells were treated with 10% trichloroacetic acid or, for autoradiography, fixed with methanol. Autoradiography was performed as described previously (Roger et al., 19831, directly in the Petri dishes. The cells were stained with toluidine blue and the proportion of labeled nuclei was determined by counting at least 1000 nuclei in each dish. Iodide transport

The cells in the Petri dishes were incubated for 2 h with ‘“‘I-labeled natrium (2 mCi/l; lo-’ mol/l) in Basal Medium Eagle (BMEI at 37°C with mercaptomethylimidazole (low3 moI/l> to block organification of iodide to proteins. They were then rapidly rinsed with BME, scraped off the dish and counted in a y-counter. The radioactivity was normalized to the cellular DNA content measured in the same cells (Roger and Dumont, 1983). CAMP measurements

Incubation of cells was stopped by scraping them into boiling water. After succinylation, the cellular CAMP concentrations were measured by radioimmunoassay (Brooker et al., 1979) using iodinated 2’0 succinyl-CAMP methyl ester as a tracer. Two-dimensional (2D) polyacrylumide gel electrophoresis

After 8 h of incubation in the presence of [3”S]methionine (100 mCi/l), the cells were rapidly rinsed at room temperature in 0.9% NaCl containing lo-” mol/l CaCl,. In order to lyse the cells and to solubilize the proteins, 240 ~1 of lysis buffer (O’FarreIl, 1975) were added to each Petri dish. Total protein radioactivity was determined as described by Lecocq et al. (1990). Two-dimensional gel electrophoresis was performed as described by Lamy et al. (1990). Proteins were separated by isoelectric focusing on cylindrical gels in the first dimension with 3.2% Servalytes pH 5-7 plus 0.8% Servalytes pH 2- 11 and according to molecular mass on sodium dodecyl sulfate, linear gradient (616%) polyac~lamide slab gets (total length of the separation gel: 18 cm) in the second dimension.

15

Results Effect of TGFPl

on DNA synthesis in human thyrocytes

Thyroid follicles were seeded with 1% serum. One day after seeding, serum was removed and the cells, which rapidly spread out and developed as a monolayer (Roger et al., 1988), were cultured for 4 days in the presence of different mitogens and of increasing concentrations of TGFPl. They were incubated for the last 24 h of the culture period with [3Hlthymidine in order to evaluate DNA synthesis. Fig. 1 (experiment 1) shows that, in these cells, TGFPl inhibited DNA synthesis induced by TSH (250 mu/l>, by fetal calf serum (10%) and by EGF (25 pg/l) plus fetal calf serum (1%). This effect was dependent on the TGFPl concentration used. It was already noticed at a concentration of TGF/31 as low as 50 rig/l while a strong inhibition occurred at a TGF/31 concentration of 5 pg/l. As shown in our previous work (Roger et al., 1988; Lamy et al., 1990) and in Fig. 1 (experiment 2) dibutyryl CAMP (10e4 mol/l) almost completely reproduced the effect of TSH (250 mu/l> on DNA synthesis. Fig. 1 (experiment 2) also shows that TGFPl (5 Fg/l) inhibited almost as strongly DNA synthesis induced by dibutyryl CAMP as that induced by TSH (67% inhibition versus 85% inhibition respectively) indicating an action downstream of CAMP formation. The extent of inhibition by TGFPl of the very weak basal DNA synthesis varied from one culture (i.e. experiment 1) to the other (i.e. experiment 2). Inhibition by TGFPl of TSH effect on DNA synthesis was observed in the six experiments performed while its inhibition of dibutyryl CAMP effect on DNA synthesis was observed in 3 experiments out of 3. Effect of TGFPI

on cell morphology

Fig. 2 shows that the retraction of the thyroid cells observed in response to TSH (250 mu/l> (Fig. 2B) was iii

EXPERIMENT

strongly inhibited by TGFpl (5 pg/l) (Fig. 2C). This inhibition occurred progressively (not shown) and was maximal after 4 days of treatment of the cells with TSH and TGFPl. The same effect of TGFj31 was observed when dibutyryl CAMP (lop4 mol/l) was used instead of TSH (not shown). By itself, TGFpl did not induce obvious morphological changes (not shown). This effect of TGFpl on cell morphology was observed in the six experiments performed. Effect of TGFPl

on iodide transport

TSH and dibutyryl CAMP stimulate iodide transport in human thyroid cells seeded and cultured in the absence of serum (Roger et al., 1988). Table 1 shows that TGFPl (5 pg/l) completely abolished the accumulation of iodide by cells cultured in the presence of TSH (250 mU/l>. This inhibitory effect of TGFPl on iodide accumulation was also observed when the cells were cultured with dibutyryl CAMP (lop4 mol/l) instead of TSH showing that TGFPl acts at a step downstream of CAMP formation. Similar results were obtained in two other experiments. Effect of TGF@l on intracellular CAMP levels

Table 2 shows that a 4-day incubation with TGFPl (5 pg/l) did not modify CAMP levels in thyroid cells cultured in control conditions, in the presence of TSH (250 mu/l> or in the presence of forskolin (lop5 mol/l). This lack of effect of TGFj31 on CAMP levels was observed in 3 experiments out of 3. Effect of TGFPl

on protein synthesis

The effect of TGFPl on protein synthesis in thyroid cells was analyzed by using two-dimensional gel electrophoresis. Proteins were labeled with [ 35S]methionine for the last 8 h of a 4-day culture in the absence or presence of TGFPl (5 pg/l> with or without TSH (250 mu/l>. In a previous paper (Lamy et al., 1990) we EXPERIMENT

1

2

J#Yri -+ 1

I j i’l

+

TSH

3+

db CAMP

CONTROL /.ig TGFWL Fig. 1. Effect of TGFpl on DNA synthesis in thyroid cells stimulated to proliferate by different mitogens. Thyroid cells were seeded in the presence of 1% serum. After 24 h they were cultured for 4 additional days in serum-free medium in the presence, in experiment 1, of 25 pg/l EGF and 1% fetal calf serum (FCS) (a), 10% fetal calf serum (X ), 250 mU/I bovine TSH (MI or control medium (+) with increasing concentrations of TGFpl and, in experiment 2, of 250 mU/l bovine TSH or of 10M4 mol/l dibutyryl CAMP with (+I or without c--j 5 pg/l TGF/31. [“H]Thymidine was present for the last 24 h of the culture period. The cells were fixed with methanol, rinsed, and processed for autoradiography. The results of each experiment are the means+ range of values of duplicate Petri dishes.

16 TABLE EFFECT

1

TABLE OF TGFpl

ON IODIDE

TRANSPORT

EFFECT

Thyroid cells were seeded and cultured without serum for 5 days. After one day, TSH (250 mU/I) or dibutyryl CAMP (10m4 mol/l) were added to the culture medium with or without TGFpl (5 yg/ll and their presence was maintained throughout the experiment. Iodide transport was measured at equilibrium on cells at day 5 as described in Methods. The results are the mean f range of values of duplicate culture dishes. Treatment

Control TSH (250 mu/I) Dibutyryl CAMP floe4 mol/l)

Iodide uptake

(cpm “‘I/ng

- TGFPl

+TGFP1.5

2.4 + 0.4 37 f 1.5

2.7 + 0.2 2.2 f 0.2

30

6.7 * 0.9

k5.0

2 OF TGFpl

Treatment

DNA) wg/l

showed that TSH enhanced, in these cells, the [35S]methionine labeling of 26 proteins (l-26) and decreased the labeling of 19 others Cl’-18’; actin). The position of these proteins in a two-dimensional gel pattern is shown in Fig. 3A. The same results were

ON CAMP LEVELS

Thyroid cells were seeded in the presence of 1% serum. 24 h later they were cultured for 4 additional days in serum-free medium in the presence of TSH (250 mu/I) or forskolin (10m5 mol/l) with or without TGFpl (5 pg/l). The results are the mean + standard deviation of the mean (SEM) of values of triplicate culture dishes.

Control TSH 250 pU/ml Forskolin lo-’ mol/l

CAMP (pmol/pg

DNA)

-TGFPl

+TGFpl5

0.5 i 0.1 17.6 * 2.0

0.6 * 0.1 20.2 + 2.3

8.1 + 1.0

7.4 + 1.o

pg/l

obtained when the cells were incubated with dibutyryl CAMP (lop4 mol/l> instead of TSH. When added to the culture medium together with TSH (250 mU/li or with dibutyryl CAMP (10m4 mol/l>, TGFj31 (5 kg/l) partly or completely inhibited a great number of the effects of these agents on protein syn-

Fig. 2. Reversal by TGFfil of cell retraction induced by TSH. Thyroid cells were cultured as described in Fig. 1 with 250 mU/l bovine TSH in the presence (Fig. 2C) or absence (Fig. 2B) of 5 pg/l TGF/31. Phase contrast microscopy was used (X 100). Fig. 2A = control.

17

acidic

Fig. 3. TGF@ mu/l) dish 8 whose whose

IEF

basic

4

basic

4

Autoradiographs of ho-dimensional gel electrophoresis of total proteins extracted from thyroid cells exposed to TSH iA) or to TSH and (B). The cells were seeded in the presence of 1% serum. One day after seeding, serum was removed and TSH (250 mU/lf or TSH (250 plus TGF/31 (5 pg/l) were added to the culture medium and maintained for 4 days. 100 &i [%]methionine was added to each Petri h before the end of the culture period. The proteins whose synthesis is modulated by TSH are identified on Fig. 3A. 1 to 26: proteins synthesis is increased by TSH; 1’ to 18’ and actin: proteins whose synthesis is decreased by the hormone. The TSH-regulated proteins synthesis is sensitive to TGF/31 are indicated in Fig. 3B. In this experiment the synthesis of proteins 17’ and 18’ was not decreased by TSH and protein 22 was not detectable. Px: thyroperoxidase; Cy: PCNA/cyclin; Tm: tropomyosin; a: actin.

thesis: 35 CAMP-regulated proteins out of 45 were affected (Fig. 3 and Table 3). Among these proteins four were identified in a previous work (Lamy et al., 1990) by their characteristic migration in two-dimensional gel ~lectrophoresis and by immunodetection. They include two cytoskeleton proteins, actin and a high molecular weight isoform of tropomyosin (protein

TABLE

IEF acidic

12’1, thyroperoxidase (protein 4), a key enzyme in the synthesis of thyroid hormones, and the proliferating cell nuclear antigen (PCNA/cyclin; protein 18) which is the auxiliary protein activating DNA polymerase S and which is required for cell proliferation. Fig. 4 illustrates the fact that TGFPl (5 pg,/l) strongly inhibited the stimulation by TSH (250 mU/l)

3

~OLYPEPTIDES WHOSE SYNTHESIS IS MODULATED BY TSH AND ~IBU~RYL CAMP AND ACTION: SUMMARY OF TWO-DIMENSIONAL GEL ELECTROPHORESIS ANALYSES

INFLUENCE

OF TGFpl

ON THIS

Thyroid cells from 3 different subjects were cultured as described in Fig. 3 in the presence of TSH (250 mu/L), dibutyryl CAMP (IO-” mol/l) and TGFPl (5 pg/l) as indicated in the table. Proteins were labeled with [%Jmethionine, separated by two-dimensional gel electrophoresis and identified by the same reference number as in Fig. 3. Polypeptide reference number

Effect of TSH or dibutyryl CAMP

Inhibition by TGFPI of TSH or dibutyryl CAMP effect

2, 9, 19,23,24,

t t

0 +

5

++

I

0

r -1

+ ++

25

4 Vx), 6, 7, 8, 10, 11,q 12, 15, 16, 17, 20, 22 J 1, 3,5, 13, 14, 18 (Cy) 21,26 1’. 7’. 10’. 11’ .5’, 6’; 8’, 9, 13’, 14’, 1.5’

‘1

2’, 3’, 4’, 12’ (Tm), 7 16’,17’,18’, actin (I 1 ( J, 1 Level of [“‘Slmethionine incorporation inhibition; + +, complete inhibition

increased

or decreased

compared

to the one observed

in control

cells. 0. no inhibition:

+ , partial

18

Mrx 10-3

TSH

Control

TGFR

TSH

+ TGFR

Fig. 4. Autoradiographic detail of [%]methionine-labeled proteins separated by two-dimensional gel electrophoresis. A characteristic section of the two-dimensional gel was chosen to illustrate the effect of TSH (250 mU/I) and TGF/31 (5 pg/l) on thyroperoxidase synthesis in thyroid cells. Thyroid cells were processed as described in Fig. 3. Arrowheads indicate the position of thyroperoxidase.

man thyroid cells of the inhibition by TGFPl of the expression of a differentiated function of these cells. Fig. 5 shows that TGFPl (5 Fug/l) not only inhibited the stimulation by TSH (250 mU/I> of the synthesis of the PCNA/cyclin but that it even inhibited the basal level of expression of this protein. The effect of TGFPl on the level of PCNA/cycIin synthesis thus correlates well with its effect on DNA synthesis (Fig. 1). By itself TGFj31 had little effect on protein synthesis. Apart from its effect on thyroperoxidase and PCNA/cyclin synthesis, it stimulated the synthesis of only one protein (protein 12’) which is an isoform of tropomyosin (Fig. 5). TSH and dibutyryl CAMP decreased the synthesis of this protein but this inhibition was relieved by the simultaneous addition of TGFPl to the culture medium (Fig. 5). The inhibition by TGFPl of the effect of TSH on actin (Fig. 3) and tropomyosin synthesis might be implicated in the reversal, by this agent, of the effect of TSH on the morphology of thyroid cells (Fig. 2). Discussion

of thyroperoxidase synthesis and shows that it even suppressed the weak expression of this protein in control cells. This constitutes the second example in hu-

x 10-3 40

-20

; Control

TSH

40

I TSH

20

-t- TGFR

Fig. 5. Autoradiographic detail of [“Slmethionine-labeled proteins separated by two-dimensional gel electrophoresis. A characteristic section of the two-dimensional gel was chosen to illustrate the effect of TSH (250 mU,/l) and TCFpl 15 pg/lf on PCNA/cyclin and tropomyt)s~n synthesis in thyroid cells. Thyroid cells were processed as described in Fig. 3. Cy: PCNA/cyclin; Tm = tropomyosin.

A growing body of evidence indicates that normal cell proliferation is controlled by a balance between positive and negative intracellular and extracellular signals. In previous studies using dog thyrocytes in primary culture, we have shown and characterized the quite distinct positive signalings of proliferation by either TSH acting through CAMP or by EGF independently of CAMP (Roger et al., 1987; Lamy et al., 1989; Dumont et al., 1989, 1992). While TSH via CAMP stimulates and maintains the functional capacity of thyroid cells (iodide transport, thyroglobulin and thyroperoxidase gene expression), EGF antagonizes these effects (Roger and Dumont, 1982; Roger et al., 1985; Pohl et al., 1990). We and others have recently extended these conclusions to human thyroid cells in culture (Roger et al., 1988; Kasai et al., 1987; Williams et al., 1988; Nagayama et al., 1989; Lamy et al., 1990). In the present study we characterized the negative effects of TGFPl, the most potent growth inhibitor for epithelial cells, on the growth and functional capacity of normal human thyrocytes in primary culture. Since TGFP mRNA and activity have been demonstrated in normal or pathological thyroid tissues or cultures (Griibeck-Loebenstein et al., 1989; Jasani et al., 1990; Morris et al., 19881, this factor might well be involved in a negative autocrine loop, as suggested in other systems (Arteaga et al., 1990; Wu et al., 1992). As shown by Tsushima and coworkers on porcine thyrocytes (Tsushima et al., 1988) and GriibeckLoebenstein et al. (1989) and Wyllie et al. (1991) on human thyrocytes we observed that TGFPl potently inhibited the stimulation of DNA synthesis by serum and EGF. In some experiments, we also detected an

19

inhibition by TGF/31 of the low basa1 level of DNA synthesis. In addition, we showed for the first time in human thyrocytes that TGF/?l almost completely inhibited the mitogenic effect of TSH. A 95% inhibition of DNA synthesis was obtained with 200 pM of TGFPl and half-maximum inhibitory effects were reached below 20 pM, which is comparable to the concentrations of TGF/31 inhibiting the TSH-dependent growth in the FRTL-5 rat thyroid cell line (Morris et al, 1988; Coiletta et al., 1989; Pang et al., 1992) and the growth of other epithelial cells (Barnard et al., 1990). It should be noted that the present results were obtained in serum-free medium, thus avoiding the possible interferences with serum factors documented in other systems (Barnard et al., 1990). No cytotoxic effect was detected even on a 4-day exposure to 5 pg/l TGF@l as judged by microscopic examination, DNA measurements, absence of inhibition of [““Slmethionine incorporation into total proteins (not shown) and absence of alteration in the pattern of protein synthesis in two-dimensional gel separation. The mechanisms of the inhibition, by TGFPl, of the TSH-dependent proliferation remains to be determined. TGFPI also markedly inhibited the stimulation of DNA synthesis by dibutyryl CAMP, which in large part reproduces the mitogenic effect of TSH (Roger et al., 1988). This restthY which contradicts a recent report on FRTL5 cells (Pang et al., 19921, indicates that, in human cells, the growth inhibition occurred downstream from the CAMP accumulation. In some but not all systems (Pietenpol et al., 1990; Chambard and Pouyssegur, 1988), the growth inhibitory effects of TGFpl have been associated with a rapid inhibition of c-myc transcription, an early event considered to be necessary in the mitogenic cascade. TGFPl may also inhibit DNA synthesis at late Gl stages, possibly through inhibition of ~34~~~’ kinase activity (Howe et al., 1991) and of phosphorylation of the plOSRB protein (Laiho et al., 1990). We observed that TGFPl also inhibits the TSH/ CAMP-dependent synthesis of PCNA/cyclin, the auxiliary protein of DNA polymerase 6 that is necessary for DNA synthesis (Liu et al., 1989), which is a late event associated with TSHstimulated mitogenesis in dog (Lamy et al., 1989) and human thyrocytes (Lamy et al., 1990). Moreover, inhibition of the CAMP effects on the synthesis of other proteins might be instrumental in the growth inhibition by TGFp 1. TGFPl has a growth inhibitory action, and is also well known as a multifunctional factor modulating various specialized functions in various systems including endocrine tissues. In thyrocytes, its effects on specialized functions and TSH responses are controversial in animal culture models and have not been studied in human cells. The present observations that TGF@l potently antagonized the stimulation by TSH and dibu-

tyryl CAMP of iodide transport and thyroperoxidase synthesis in human thyrocytes are in agreement with reports by Tsushima et al. (1988) on porcine cells and Colletta et al. (1989) and Pang et ai. (1992) on rat thyroid FRTL-5 cells, but they strikingly contrast with the observation by Morris (1988) of a stimulation, by TGFPl, of iodide transport in FRTL-5 cells. However, a 4-day exposure to TGFPl did not affect the stimulation by TSH of CAMP accumuIation in human thyrocytes, which is in accordance with Tsushima’s report on porcine cells and that of Colletta on FRTL-5 cells, but contradicts data of Winter et al. (1989) on porcine cells and Morris et al. (1988) on FRTL-5 cells. TGFPl, like EGF (Roger et al., 1988; Lamy et al., 1990) thus appears as a potent local antagonist of TSH effects on the functional capacity of human thyrocytes, acting distal to CAMP accumulation. As in FRTL-5 cells (Garbi et al., 19901, the morphological effects of TSH and dibutyryl CAMP were progressively inhibited by TGFPl, although TGF/31 alone did not obviously influence the morphology of human thyrocytes at variance with its effect on FRTL-5 cells (Garbi et al., 1990). The morphological effects of TSH and CAMP are associated with a disorganization of microfilament bundIes and a decrease in the synthesis of microfilament proteins including actin and high M, isoforms of tropomyosin (Westermark and Porter, 1982; Roger et al., 1989; Lamy et al., 1990). By contrast, TGFPl induces stabilization of microfilament bundles in FRTL-5 cells (Garbi et al., 1990) and, in the present experiments, it stimulated the synthesis of a high M, tropomyosin isoform f T, 3) and it counteracted the inhibitory effects of TSH and CAMP on the synthesis of this tropomyosin isoform and of actin. Whether the reversal by TGFPl of TSH effects on morpholo~ and cytoskeleton proteins is related to the abundantly documented effects of this agent on the extracellular matrix composition and organization as in other cell types (Massagu~, 1990) is unknown. While TGFPl per se had only very little effect on the pattern of gene expression as revealed by the two-dimensional gel electrophoresis of neosynthesized proteins, it was unexpected to find that it reversed a large majority of the effects of TSH and dibutyryl CAMP on the synthesis of proteins. TSH, through CAMP, positively or negatively modulated the synthesis of at least 45 proteins in human thyrocytes (Lamy et al., 19901, and TGFPl antagonized the CAMP effect on 35 of these proteins. On the combined basis of the effects on growth, differentiation, morpholo~ and protein synthesis, it can therefore be concluded that TGF/31 appears as an almost general antagonist of the TSH and CAMP cascade in the human thyroid. Although TGFPl is known to have various effects on gene expression in different systems [such as induction of c-jun and jun B (Pertovaara et al., 19891, autoinduc-

20

tion of TGFPl (Kim et al., 19901, inhibition of growth factor activation of myc and KC genes (Coffey et al., 1988) and of some metalloprotease genes (Kerr et al., 199011, to our knowledge, there is no precedent for such a general inhibitory effect. According to proposals that the spatial organization and morphology of cells may influence their growth and differentiation characteristics, as well as gene expression (Ben-Ze’ev, 19891, we have suggested that the positive or negative effects of TSH via CAMP on the synthesis of different proteins (28 out of 4.5) may depend in part on the morphological changes that CAMP induces (Lamy et al., 1990). The inhibitory action of TGFj31 on the modulation, by CAMP, of the synthesis of many proteins might thus reflect its reversion of the morphological action of CAMP. Nevertheless, TGFPl also inhibited the TSH effects on the synthesis of several proteins (12 out of 35 that are affected by TGFPll, including PCNA and thyroperoxidase, which are not responsive to morphological changes (Lamy et al., 1990). The wide range of inhibition, by TGFPl, of CAMP responses should also be compared to the situation we found previously with EGF, which appears as another antagonist of TSH responses in human thyrocytes (Lamy et al., 1990). Interestingly, inhibitory effects of TGFPl and EGF overlap on 19 proteins. However, TGFPl potently repressed the TSH effects on the synthesis of a few proteins that are unaffected by EGF, and conversely. Unlike TGFPl, EGF weakly influenced TSH effects on growth (Lamy et al., 1990). Whether the unidentified proteins whose synthesis is similarly affected by TGFj31 and EGF are related to differentiation and morphology is unknown. The present study therefore illustrates the fact that the stimulatory effects of TSH on human thyrocytes, not only on growth but even on their functional capacity, cannot be considered outside the context of local modulators such as TGFPl (or EGF). Acknowledgements We thank Dr J. Van Sande and C. Massart for CAMP determinations and C. Doudelet for technical assistance with the two-dimensional gel electrophoresis. This work was supported by grants from the ‘Minis&e de la Politique Scientifique’ (Sciences de la Vie), the Cancer Research Fund of the ‘Caisse G&n&ale d’Epargne et de Retraite’, the ‘Association contre le Cancer’ and Euratom (contract BlO-C-36081-B to Prof. J.E. Dumont). P.P. Roger is a Research Associate of the National Fund for Scientific Research (Belgium). References Arteaga, CL., Coffey Jr, R.S., Dugger, T.C., McCutchen, C.M., Moses, H.L. and Lyons, R.M. (1990) Cell Growth Differ. 1, 3677374.

Barnard, J.A., Lyons, R.M. and Moses, H.L. (1990) Biochim. Biophys Acta 1032, 79-87. Ben-Ze’ev, A. (1989) in Cell Shape: Determinants, Regulation and Regulatory Role, pp. 95- 119. Academic Press, New York. Brooker, G., Harper, J.F., Teraski, W.L. and Moylan, R.D. (1979) Adv. Cyclic Nucleotide Res. 10, 1-34. Chambard. J.C. and Pouyssegur, J. (1988) J. Cell. Physiol. 135, 101-107. Coffey, R.J., Bascom, CC., Sipes. N.J., Graves-Deal, R., Weissman, B.E. and Moses, H.L. (1988) Mol. Cell. Biol. 8, 3088-3093. Colletta, G., Cirafici, A.M. and DiCarlo, A. (1989) Cancer Res. 49, 3457-3462. Dumont, J.E., Jauniaux, J.C. and Roger. P.P. (1989) Trends Biochem. Sci. 14, 67-71. Dumont, J.E., Lamy, F., Roger, P. and Maenhaut, C. (1992) Physiol. Rev. 72. 6677697. Garbi, C., Colletta, G., Cirafici. A.M., Marchisio, P.C. and Nitsch, L. (1990) Eur. J. Cell. Biol. 53. 281-289. Griibeck-Loebenstein. B.. Buchan, G., Sadeghi, R., Kissonerghis, M.. Londei, M., Turner, M., Pirich, K., Roka, R., Niederle, B., Kassal, H., Waldhausl, W. and Feldmann, M. (1989) J. Clin. Invest. 83, 764-770. Howe, P.M., Draetta, G. and Leof, E.B. (1991) Mol. Cell. Biol. II, 11x5-1194. Jasani, B., Wyllie, F.S., Wright, P.A., Lemoine, N.R.. Williams, E.D. and Wynford-Thomas, D. (1990) Growth Factors 2, 149-155. Kasai, K., Hiraiwa, M., Sijuki, Y.. Emoto, T., Banba, N., Nakamura, T. and Shimoda, S. (1987) Acta Endocrinol. 114, 396-402. Kerr, L.D.. Miller, D.B. and Matrisian, L.M. (1990) Cell 61, 267-278. Kim, S.J.. Angel, P., Lafyatis, R., Hattori, K., Kim, K.Y., Sporn. M.B., Karin, M. and Roberts, A.B. (1990) Mol. Cell. Biol. IO. 1492- 1497. Laiho, M., De Caprio, J.A., Ludlow, J.W., Livingston, D.M. and Massague, J. (1990) Cell 62, 175-185. Lamy, F., Roger, P., Lecocq. R. and Dumont, J.E. (1989) J. Cell. Physiol. 138, 568-57X. Lamy, F., Taton, M., Dumont, J.E. and Roger, P.P. (1990) Mol. Cell. Endocrinol. 73, 1955209. Lecocq. R., Lamy, F. and Dumont, J.E. (1990) Electrophoresis 11, 200-212. Liu, Y.C., Marraccino, R.L., Keng, P.C., Bambara, R.A., Lord, E.M.. Chou, W.G. and Zain, S.B. (1989) Biochemistry 2X, 2967-2974. Massague, J. (1990) Ann. Rev. Biol. 6, 597-641. Morris, J.C., Ranganathan, G., Hay, I.D., Nelson, R.E. and Jiang, N.S. (1988) Endocrinology 123, 1385-1394. Nagayama, Y., Yomashita, S., Hirayu, M., Izumi, M., Uga, T., Ishikawa, N., Ito, K. and Nagataki, S. (1989) J. Clin. Endocrinol. Metab. 68, 1155-l 159. O’Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021. Pang, X., Park, M. and Hershman, J.M. (1992) Endocrinology 131, 45-50. Pertovaara, L., Sistonen, L., Bos, T.J.. Vogt, P.K., Keski-Oja, J. and Alitalo. K. (1989) Mol. Cell. Biol. 9, 1255-1262. Pietenpol, J.A., Holt, J.T., Stein, R.W. and Moses, H.L. (1990) Proc. Natl. Acad. Sci. USA 87, 3758-3762. Pohl, V., Roger, P.P.. Christophe, D., Pattyn, G., Vassart, G. and Dumont, J.E. (1990) J. Cell Biol. 111, 663-672. Roberts, A.B. and Sporn, M.B. (1988) Adv. Cancer Res. 51, 107-145. Roger, P.P. and Dumont, J.E. (1982) FEBS Lett. 144, 209-212. Roger, P.P. and Dumont, J.E. (1983) J. Endocrinol. 96, 241-249. Roger, P.P., Rickaert, F., Lamy, F., Authelet, M. and Dumont, J.E. (1989) Exp. Cell Res. 182, 1-13. Roger, P.P., Setvais, P. and Dumont, J.E. (1987) J. Cell. Physiol. 130, 58-67. Roger, P.P., Servais, P. and Dumont, J.E. (1983) FEBS Lett. 157. 323-329. Roger, P.P., Taton, M., Van Sande, J. and Dumont. J.E. (1988) J. Clin. Endocrinol. Metab. 66, 115X-l 165.

21 Roger, P.P., Van Heuverswyn, B., Lambert, C., Reuse, S., Vassart, G. and Dumont, J.E. (1985) Eur. J. Biochem. 152, 239-245. Tsushima, T., Arai, M., Saji, M., Ohba, Y., Murakami, H., Ohmura, E., Sato, K. and Shizume, K. (1988) Endocrinology 123, 11871193. Westermark, B. and Porter, K.R. (1982) J. Cell Biol. 94, 42-50. Williams, D.W., Williams, E.D. and Wynford-Thomas, D. (1988) Br. J. Cancer 57, 535-539.

Winter, H., Bechtner, G. and Greil, W. (1989) Ann. Endocrinol. (Paris) 50, 89 (abstract). Wu, S., Theodorescu, D., Kerbel, R.S., Willson, K.V., Mulder, K.M., Humphrey, L.E. and Brattain, M.G. (1992) J. Cell Biol. 116, 187-196. Wyllie, F.S., Dawson, T., Bond, Y.A., Goretzki, P., Game, S., Prime, S. and Wynford-Thomas, D. (1991) Mol. Cell. Endocrinol. 76, 13-21.